-
-
Notifications
You must be signed in to change notification settings - Fork 256
Issues: dask/dask-ml
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Author
Label
Projects
Milestones
Assignee
Sort
Issues list
KMeans Optimization: Incorporating Weights into Re-Clustering Process
#1001
opened Oct 3, 2024 by
ChiaTrama
Area under the receiving operating characteristic curve (AUROC) calculation.
#977
opened Sep 5, 2023 by
stephenpardy
For a single record data frame train_test_split() sometimes assigns this single record to test set.
#975
opened Jun 29, 2023 by
KWiecko
Incremental does not handle dask arrays of ndim>2 in estimator training
#972
opened Jun 21, 2023 by
azagajewski
sklearn handles text labels differently than ml_dask on OneHotEncoding
#964
opened Feb 8, 2023 by
brian-methodical
HashingVectorizer behaves differently from FeatureHasher
#963
opened Feb 8, 2023 by
brian-methodical
Add backward compatibility for supported version of scikit-learn
#961
opened Jan 18, 2023 by
mmccarty
LabelEncoder doesn't handle missing values in *dask* series of strings
#954
opened Dec 15, 2022 by
phobson
LabelEncoder raises errors with string and string[pyarrow] types
#953
opened Dec 15, 2022 by
phobson
Default datatype when using CountVectorizer and HashingVectorizer should be sparse COO
needs triage
#952
opened Nov 15, 2022 by
alejohz
Previous Next
ProTip!
Type g p on any issue or pull request to go back to the pull request listing page.